[llvm] [NVPTX] Add intrinsics for narrow-fp to bf16 conversions (PR #191376)

[Rajat Bajpai via llvm-commits]

================
@@ -729,6 +743,16 @@ let Predicates = [callSubtarget<"hasS2F6X2ConversionSupport">] in {
                                       "cvt${mode:base}${mode:relu}.f16x2." # type>;
   }
   
+  multiclass CVT_F6x2_TO_BF16x2<string F6Name> {
----------------
rajatbajpai wrote:

This could be merged with above `fp6x2 -> f16x2 (unscaled)` under a single multiclass.

https://github.com/llvm/llvm-project/pull/191376